On Fri, Jul 03, 2026 at 02:20:08AM +0500, Kirill Reshke wrote:
> I was debugging the GIN index issue today and noticed my postgresql
> session which I used to run amcheck eats too much memory. PFA fix.
Hmm.. Still need to do some measurements with a large GIN index to
get an idea, but a quick lookup at the code points that you are right.
This could stack quite a lot in a single pass even if we have a
dedicated memory context while in gin_check_parent_keys_consistency().
Will measure and check.
--
Michael